Zhang, Peng, et al. “FaNe: Towards Fine-Grained Cross-Modal Contrast With False-Negative Reduction and Text-Conditioned Sparse Attention”. Proceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 15, Mar. 2026, pp. 12681-9, doi:10.1609/aaai.v40i15.38264.